Tobacco Smoking and Reduced Parkinson’s Risk

Various Experimental Answers to the Epidemiological Question


By Anthony Gregory


Written for Molecular Cell Biology at UC Berkeley, taught by Dr. David Presti, Ph.D.

May 6, 2003

An Epidemiological Introduction

The available epidemiological evidence indicates a negative correlation between tobacco use and likelihood in acquiring Parkinson’s Disease. Since a 1959 study, which first exposed this phenomenon, the data suggesting the correlation have been widely discussed and fairly controversial. (Allam et al. 2002) According to a case-control study from 2000 that reviewed epidemiological studies, almost all of them indicated that the decrease in risk was significant among smokers. At least some studies yield results of smokers being nearly 50% less likely than their non-smoking counterparts to develop the disease.  (Benedetti et al. 2000). Others reveal even more compelling numbers. In one study current smokers have a 60% reduced risk. (Martyn and Gale, 2003) The negative correlation has been reproducible. (Jeyarasasingham et al. 2002)

However, a number of other studies have shown no significant difference in nicotine exposure between those with and without Parkinson’s. This has led some scientists to question the epidemiological evidence, arguing that other factors entered into data collection. A meta-analysis from 2002 extensively reviewed experimental literature on the subject, specifically looking for disparities between Parkinson’s Disease risk for subjects, smokers and non-smokers, with and without positive Parkinson’s disease family histories. Arguing that recent hypotheses only found a demonstrated reduction in Parkinson’s Disease likelihood for those with negative family histories, the scientists aimed to examine what they considered a neglected factor: the effect of smoking tobacco on those with positive family histories. They came to the conclusion that the answers to the mysterious and controversial link between smoking and Parkinson’s lay in this stratification between subjects with positive and negative family histories. (Allam et al. 2002)

Another study from a couple years earlier also came to conclusions that the epidemiological evidence was not all it was cracked up to be. Invoking other epidemiological studies that showed similar correlations between alcohol or coffee and Parkinson’s, the scientists said they suspected a “secondary association” between such chemical intake and the disease. Perhaps nicotine use – or alcohol or coffee use – was indicative of other environmental or demographic factors in the test subjects, which would in turn be the underlying explanation of their higher or lower susceptibility to Parkinson’s.  They hypothesized that rather than nicotine having a neuroprotective effect – or other biological advantage that would reduce Parkinson’s risk ­– people who were predisposed to Parkinson’s were simply more likely to avoid smoking (or alcohol or coffee), because they possess “a premorbid personality manifesting early in life, which leads them to avoid the use of substances that cause dependence, are negatively sanctioned by society, or may jeopardize health.” (Benedetti et al. 2000)

I find there are some problems with this hypothesis and this study. It seems as though they overlook ways in which their assumptions may invalidate their own data. They assume that the lower risk for alcohol and coffee users is a reason to suspect a “secondary association.” But they do not see how such an association can exist between smokers and drinkers. Perhaps there is a real negative correlation between nicotine use and Parkinson’s, and a real positive correlation between smokers and drinkers, leading to the secondary association between drinkers and Parkinson’s. At any rate, they do not adequately address such possibilities, which seem more intuitively plausible than some correlation between being predisposed to Parkinson’s and being adverse to drinking coffee or beer because of their perceived health dangers or social stigma.

While discussions over the validity of the epidemiological evidence may persist, the value of such evidence has been sufficient in the last few decades to make scientists wonder about and attempt to explain the suspected correlation. Regardless of whether the epidemiological studies were biased or flawed, the curiosity they have sparked has prompted a number of physiological studies into biological mechanisms thought to be at the root of the correlation between tobacco and a decrease in risk for Parkinson’s. For a while many considered the possibility that nicotine helped the activation of dopamine, but this would only yield a temporary resistance to Parkinsonian symptoms. (Ferger et al. 1998) There seem to exist at least several other major postulates, with corresponding experiments, behind the physiological causes of the phenomenon. These include the quite recently considered possibility that nicotine protects mitochondria, the hypothesis that nicotine shields neurons from oxidative damage and general neurotoxicity, and the notion that other chemicals in tobacco interfere with monoamineoxidase B in platelets. Because of the varying mechanisms under scrutiny and the limited degree of relevance experiments on different properties have for each other, it is valuable to review these mechanisms and corresponding experiments one at a time.

Shielding Mitochondria

In a very recent study in 2003, scientists explored the possibility that one of the protective properties of nicotine has to do with its preservation of mitochondrial functions. The study seems important because it is among the most recent developments on the issue of nicotine’s effect on Parkinson’s, and while mitochondria has for years been suspected as important in this matter, it has received very little specific attention. The authors of the study cite sources that have shown a correspondence between mitochondrial failure and neuron death from a variety of neuordegenerative diseases, including Parkinson’s and Alzheimer’s. The study explains some of the specifics of this correlation. Alzheimer’s disease sometimes involves mutated mitochondrial genes and mitochondrial defections can lead to pathogenesis of Alzheimer’s. Mitochonrial failure might also be at the root of oxidative damage to portions of the substantia nigra in Parkinson’s patients. (Cormier et al. 2003)

Drawing upon the epidemiological suggestion that nicotine helps combat these neurological diseases, as well as the importance of mitochondria in these pathologies, the study attempts to find out if nicotine has the same effects on mitochondria in vivo experiments as it did in vitro ones, as well as carry out an analysis of nicotine on mitochondria. For the test, they used rats. (Cormier et al. 2003)

The experiment went as follows:

Scientists made incisions in rats and pumped nicotine into some of them and only saline into others for control purposes. The rats were caged for seven or fourteen days at which point they were killed and their forebrains removed so mitochondria could be isolated. To do this, scientists homogenized the forebrains and used differential centrifugation. Using an isolation buffer technique, they isolated the mitochondria. The scientists then proceeded to use incubation and substrate measurements to determine levels of the mitochondria’s oxygen consumption. Using related methods, they also tested for mitochondrial superoxide anion generation and mitochondrial membrane anisotropy. (Cormier et al. 2003)

They found that the nicotine considerably reduced how much oxygen the mitochondria consumed. The results were consistent and dose dependent (the higher the nicotine concentrations administered, the less oxidation on the part of the mitochondria). High doses of nicotine brought oxygen consumption down to less than 70% the levels seen in mitochondria unexposed to nicotine, and their findings for in vitro experimentations had very similar results. Likewise, similarly beneficial effects of nicotine were found in its ability to protect mitochondria from rotenone-related damage, including its tendency considerably to offset the membrane anisotrophy  that is characteristic of rotenone exposure. This has important implications because rotenone is a neurotoxin suspected to be an environmental cause of Parkinson’s. Nicotine also was shown to decrease superoxide anon generation. (Cormier et al. 2003)

Interestingly enough, the scientists who studied the effects of nicotine on mitochondria concluded that such effects were independent of action having to do with nicotinic receptors. Nicotine’s affinity was separate from – and higher than – its propensity toward receptors. (Such an independence is further validated by the lack of detected effects of nicotinic agonists and antagonists on mitochondria.) in vitro experiment, having similar results as the in vivo one, also helped to demonstrate that it was nicotine, and not its metabolites, responsible for the effect on mitochondria. This study presents a compelling case for the importance of mitochondria in nicotine’s apparent relationship with Parkinson’s, but it suggests that there may be other mechanisms at work, perhaps having to do with oxidation or involving nicotinic receptors.  (Cormier et al. 2003)

Oxidative Damage Prevention Unrelated to Nicotinic Receptors

Before the recent study focusing on the effects of nicotine on mitochondria, much of the scholarship concerning the drug’s beneficial effects in respect to Parkinson’s circulated around oxidation and the neuroprotective effects of nicotine against oxidation. This oxidation comes in a couple of forms, including free radical damage and damage from MPTP in the basal ganglia via monoamineoxidase B. MPTP’s significance for Parkinson’s has been studied in mice ever since its first discovered propensity to cause mice (and humans) to acquire Parkinson’s. (Maggio 1998) In short, the body uses MAO B to metabolize MPTP into MPP+, which is pulled into the substantia nigra’s domaminergic system, reducing mitochondrial respirtation and bringing about free radicals and cell death for neurons. (Quick, 2001) A lot of research has come out in the last several years to shed light on the relationship between nicotine – epidemologically demonstrated (to the satisfaction of some, anyway) to have a negative correlation to Parkinson’s – and MPTP. The research has yielded exciting results.

Several of the experiments tested the effects of nicotine on MPTP-induced damage in mice. In one study, administration of nicotine demonstrably prevented the parkinsonism that experimenters successfully induced in other mice by giving them MPTP but no nicotine. The study found that nicotine inhibited DCC, an agent that enhances MPTP toxicity by increasing dopamine depletion in the striatem. It found, however, that nicotine could not protect against MPTP in itself. In this respect – acting on DCC to reduce MPTP damage –  nicotine was found to act remarkably like the chemical (+)-MK-801. The scientists conducting the study conclude that the same mechanism might be involved when nicotine is administered and when (+)-MK-801 is. They further conclude that the activity involved is not related to nicotine binding at the NMDA receptor, because whereas the (+) isomer of nicotine has a higher affinity for the receptor than the (-) isomer, the latter has a much more prevalent effect in subduing DCC than the former, which has almost none at all. The study also raises questions as to the importance of growth factors in the success of domaminergic systems. Their experiments showed that nicotine enriched levels of striatal FGF-2, a receptor usually depleted in the substantia nigra of Parkinson’s patients. (Maggio et al. 1998)

Another study also looked at nicotine’s protection of the substantia nigra and striatum from MPTP-related damage in mice, focusing on mechanisms unrelated to nicotinic receptors. To do this, the experimenters also looked at how nicotine regulated striatal quantities of MPP+ in mice. Like their counterparts in the previously mentioned experiment, the mice were given nicotine and MPTP, killed, and their brains were centrifuged. In this case, however, they only lived hours after the injections and before they were sacrificed to science. Those mice that received nicotine had about a third the level of MPP+ in their striata as those who received none. It was found additionally that nicotine, at whatever concentrations, had no effect on the activity of MAO A or MAO B – indicating that the protective effects of nicotine had nothing to do with the conversion of MPTP into MPP+, a process triggered by MAO. (Quik et al. 2001)

The study gave some possible explanations for the mechanisms at work in the tobacco use/Parkinson’s matter. Perhaps the increased dopamine from nicotine intake causes an incidental increase in MPP+ release from nerve terminals and consequent decrease in MPP+ in the striatum. Perhaps nicotine reduces the presence of other neurotoxins in the striatum somehow. Perhaps the nicotinic receptors are involved somehow, or some chemicals in tobacco are having an effect on MAO – other than nicotine, shown to have none. (Quik et al. 2001)

Nicotinic Receptors as a Focus

Some have researched in particular the possibility of nicotine having a direct effect on neurotoxicity through activation of nicotinic receptors. Certain receptors have been implicated in neuroprotection, including a3 through a7 and b2  through  b4. Scientists have found especially interesting a6 and b3 because of their high prevalence in the substantia nigra and scarcity elsewhere. a6 has been found in particular to mediate movement in mice. In one experiment last year, scientists used a culture test in attempts to determine the nicotinic receptors involved in regulating MPTP-related damage to the substantia nigra. (Jeyarasasingham et al. 2002) åAads fsj

In the experiment, scientists sacrificed mice and centrifuged and cleaned the brains. The cells were prepared in a culture, which was incubated, and then radioactivity counting was used to detect the cells. On the third day, scientists administered differing potencies of nicotine into the cultures. Through antibody detection, they counted neurons and measured dopamine uptake. (Jeyarasasingham et al. 2002)

They came to a number of useful conclusions. First, and consistent with most of the other research, they found that nicotine had a profound effect on damage from MPTP. Secondly, they found that this desired effect did not come about because nicotine was blocking MPP+ uptake into dopamaminergic neurons: levels of [3H] dopamine uptake were unaffected by nicotine exposure. Third, and most important: nicotinic receptors are involved in nicotine’s nuero-protective effects. D-tubocurarine (10-4M), a non-selective antagonist at nicotinic receptors, had the effect of eliminating the neuroprotective properties of nicotine. (Interestingly, when [125I]a-bungarotoxin – a selective antagonist that binds to the a7 receptor – was used, the neuroprotective qualities of nicotine remained intact: presumably, non-a7 receptors are at work.) The study falls short of claiming to know exactly which receptors are operating; the scientists assume a3, a5, a6, b3, and b4 are involved – perhaps in a dopaminergic or GABAergic context – but they do not claim to know exactly how or why. The study also indicates that there is more going on than just the nicotinic receptor action – a sentiment that leaves room for the other researchers’ work. (Jeyarasasingham et al. 2002)


Other Chemicals in Tobacco

As Quik et al said, receptors might be involved in the neuroprotective properties of tobacco. The study also opened up the possibility ingredients other than nicotine in tobacco were involved in neuroprotection. It in fact suggested that some other chemical could be reducing MAOs in the brain, which would both explain their findings that nicotine had no effect on MAOs and the epidemiological tendency of smokers to have less MAO. (Quik et al. 2001)

Too much MAO B can cause oxidation of dopamine and an excess of free radicals. These can cause mitochondrial damage and other damage when reacting with amino acids, lipids and DNA. Active smokers have a 40% smaller binding affinity of the MAO B’s tracer substance. While nicotine – along with other tobacco chemicals, including thiocyanate and hydrazine – do not inhibit MAO B, the evidence indicates that something in tobacco does. Norharman, a substance in tobacco, is an endogenous MAOI. In one study, scientists examined if smoking produces amounts of norharman sufficient to inhibit MAO. (Rommelspacher et al. 2002)

In this experiment smokers and nonsmokers smoked a cigarettes and levels of norharman and its derivative harman in different systems of the body, including platelets and plasma. Among the fascinating discoveries was that though smokers got higher doses of harman and norharman – probably from inhaling deeper due to habit – the two groups had nearly identical turnover rates. And in spite of all of this, nonsmokers saw more of an effect of tobacco on their MAO activity. This indicates that perhaps cumulatively, smokers have blocked certain binding sites, which therefore require more tobacco smoking to bind to than do the virginal sites in nonsmokers. The study concluded that smoking introduced enough norharman to the system to interfere with MAO B in the platelets, as well as enough harman to inhibit MAO A. So whereas some chemicals, like deprenyl, primarily affect only one MAO isoform thereby exposing dopaminergic neurons to neurodegeneration, components of tobacco smoke work on both fronts can effectively reduce the damage of both types of MAO. By reducing this damage, chemicals other than nicotine in tobacco can detectably reduce the risk of Parkinson’s.  (Rommelspacher et al. 2002)


It looks like the mechanisms at work in reducing Parkinson’s prominence among smokers may be varied and complicated. Other studies have examined many of the same biological systems as have the previously discussed experiments, and scientists will likely consider new possibilities as well. As of now, the main explanations are varied: Some property of nicotine is preventing oxidative damage especially in the striatum and substantia nigra and independent of receptor binding; nicotine is in fact acting via the receptors to prevent damage against such neurotoxins as MPTP; as researched most recently,  nicotine has a particularly beneficial effect on mitochondria, including on its intake of oxygen; other chemicals in tobacco other than nicotine may be reducing the effectiveness of MAOB, whose over-activity can prove neurotoxic. These are all plausible explanations that can in most cases overlap. In some cases, however, the studies have conclusions difficult to reconcile. (e.g. Maggio et al. shows with nicotinic agonists and antagonists that receptors are not the main mechanisms at work; Quik et al. suggests that receptors may be part of the answer; Jeyarasasingham et al. outright demonstrates that nicotinic receptors are involved.)

What seems clear, however, is that something is happening with the components of tobacco and Parkinson’s. Unfortunately, even this broad statement relies on contested evidence. The epidemiological studies have indicated a correlation, but have faced challenges from some scientific circles. The physiological evidence seems to verify the correlation, but even there some discrepancies exist. For example, in one test in the late nineties, nicotine was actually found in larger to doses to increase the effects of MPTP. (Ferger et al. 1998) The studies at hand have revealed some valuable insights. Even one of the most poisonous chemicals to which humans are frequently exposed might hold some therapeutic value.  To unlock such potentially revolutionary mysteries, researchers must continue their work to test the credibility of the currently understood hypotheses, as well as to explore new ideas. It’s amazing that even on a chemical as socially and legally accepted as nicotine such uncertainty exists after so much testing. And yet, so much has been learned. It would be hard to predict how much scientific knowledge would blossom in civilization if scientists could access and study freely all the poisons in our world.


Allam, M.F. et al. Parkinson’s disease, smoking and family history: meta-analysis. European Journal of Neurology 10: 59-62 (2003).

Benedetti, M.D. et al. Smoking, alcohol, and coffee consumption preceding Parkinson’s disease: A case-control study. Neurology 55: 1350-1358 (2000).

Cormier, A. et al. Nicotine protects rat brain mitochondria against experimental injuries. Neuropharmacology Volume 44, Issue 5: 642-652 (2003).

Ferger, Boris et al. Effects of nicotine on hydroxyl free radical formation in vitro and on MPTP-induced neurotoxicity in vivo. Naunyn-Schmiedeberg’s Arch Pharmacol 358: 351-359 (1998).

Jeyarasasingham, G. et al. Simulation of non-a7 nicotinic receptors partially protects dopaminergic neurons from 1-methyl-4-phenylpyridinium-induced toxicity in culture. Neuroscience Vol. 109, No. 2: 275-285 (2002).

Maggio, Roberto et al. Nicotine Prevents Experimental Parkinsonism in Rodents and Induces Striatal Increase of Neurotrophic Factors. Journal of Neurochemistry Vol. 71, No. 6: 2439-2446 (1998).

Martyn, Christopher and Chris Gale. Tobacco, coffee, and Parkinson’s disease: Caffeine and nicotine may improve the health of dopaminergic systems. BMJ 326: 561-562 (2003).

Rommelspacher, Hans et al. The levels of norharman are high enough after smoking to affect monoamineoxidase B in platelets. European Journal of Pharmacology 441: 115-125 (2002).